NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Committee on the Consequences of Uninsurance. Coverage Matters: Insurance and Health Care. Washington (DC): National Academies Press (US); 2001.

Cover of Coverage Matters

Coverage Matters: Insurance and Health Care.

Show details

1Why Health Insurance Matters

The Institute of Medicine (IOM) Committee on the Consequences of Uninsurance launches an extended examination of evidence that addresses the importance of health insurance coverage with the publication of this report. Coverage Matters is the first in a series of six reports that will be issued over the next two years documenting the reality and consequences of having an estimated 40 million people in the United States without health insurance coverage. These reports will examine the implications of lacking health insurance for those without it, for their families, for communities in which a substantial number of people are uninsured, and for this country as a whole.

The Committee will look at whether, where, and how the health and financial burdens of having a large uninsured population are felt, taking a broad perspective and a multidisciplinary approach to these questions. To a great extent, the costs and consequences of uninsured and unstably insured populations are hidden and difficult to measure. Financial costs incurred by those without coverage may be covered by payments for the health care of those with insurance or paid by charities and taxpayers, and the health effects may be absorbed by families in the form of diminished physical and psychological well-being, productivity, and income.

The goal of this series of studies is to refocus policy attention on a longstanding problem. Following the longest economic expansion in American history, in 1999, an estimated one out of every six Americans—32 million adults under the age of 65 and more than 10 million children—remains uninsured (Mills, 2000). A better understanding of the consequences of existing policies and health care financing arrangements should reinvigorate discussions of the issue of coverage and better equip us to design and evaluate policy initiatives and proposed reforms intended to address this problem.

The Committee's charge is to communicate to the public and policy makers analytical findings about the meaning of a large uninsured population for individuals, families, and their communities, as well as for society as a whole. Its reports should contribute to the public debate about insurance reforms and health care financing by assessing the theoretical and empirical research in health services, medicine, epidemiology, and economics that bears on the effects of lacking health insurance. It is not within the scope of this project to develop or advocate for a specific set of reforms or policies.

The goal of this first report is to provide background for the findings and conclusions that the Committee will present in subsequent reports about the consequences of uninsurance by including common definitions and an overview of the dynamics of health insurance coverage. This report addresses the extent to which Americans are without coverage, identifies social, economic, and policy factors that contribute to the existence and persistence of an uninsured population in the United States, and reports the probability for members of various population groups of being uninsured. In addition, it introduces a conceptual framework that models how health insurance affects access to health care services and, through such access, affects health and economic well-being. This framework will guide the analysis in succeeding reports in the series and will be modified to address each report's set of topics.


The first step in identifying and measuring the consequences of being without health insurance and of high uninsured rates at the community level is to recognize that the purposes and constituencies served by health insurance are multiple and distinct. These purposes include promoting health, obtaining health care for individuals and families, and protecting people financially from exceptional health care costs. Health insurance pools the risks and resources of a large group of people so that each is protected from financially disruptive medical expenses resulting from an illness, accident, or disability. In addition to serving the typical functions of risk insurance, health insurance has developed as a mechanism for financing or pre-paying a variety of health care benefits, including routine preventive services, whose use is neither rare nor unexpected. Despite the fact that a large proportion of persons with health insurance make claims against their coverage every year, health care spending, and thus health insurance payouts, remain concentrated among a relatively small number of claimants, who incur high costs for serious conditions. Ten percent of the population accounts for 70 percent of health care expenditures, a correlation that has remained constant over the past three decades (Berk and Monheit, 2001). Thus health insurance continues to serve the function of spreading risk even as it increasingly finances routine care. From the perspective of health care providers, insurance carried by their patients helps secure a revenue stream, and communities benefit from financially viable and stable health care practitioners and institutions.

Employers offer health benefits both to attract and retain workers and to maintain a productive workforce. Government provides health insurance to populations whom the private market may not serve effectively, such as disabled and elderly persons, and populations whose access to health care is socially valued, such as children and pregnant women.

The ultimate ends of health insurance coverage for the individual and communities, including workplace communities of employees and employers, are improved health outcomes and quality of life. Attributing success in achieving these goals to health insurance alone presents a challenge because isolating the relative contribution of different determinants of individual and population health requires a complex analysis. Over the past quarter of a century, the importance of health insurance has grown, as clinical medicine has become increasingly sophisticated, technological advances have become more commonplace, and the range of therapeutic interventions (and their costs) has expanded rapidly. As a society, we invest heavily in health insurance through direct personal expenditures, forgone wages, and tax policy. Health insurance in the United States has developed as a common but not universal component of the employment contract. Employees rank health insurance first by far in importance among all the benefits offered in the workplace (Salisbury, 2001). Although there have been sizable investments of personal and public funds to provide health insurance, many people still have no coverage.


Despite extensive reporting of survey findings and health care research results, the general public remains confused and misinformed about Americans without health insurance and the implications of lacking coverage. This section presents basic information about health insurance and who lacks it in the context of several pervasive popular myths. Without question, the complexity of American health care financing mechanisms and the wealth of sources of information add to the public's confusion and skepticism about health insurance statistics and their interpretation. This report and those that will follow aim to distill and present in readily understandable terms the extensive research that bears on questions of health insurance coverage and its importance.

Myth: Uninsured people get health care when they really need it. Fifty-seven percent of Americans polled in 1999 believed that those without health insurance are “able to get the care they need from doctors and hospitals” (Blendon et al., 1999, p.207). In 1993, when national attention was focused on the problems of the uninsured and on pending health care legislation, just 43 percent of those polled held this belief (Blendon et al., 1999).

Reality: The uninsured are much more likely to forgo needed care (Schoen and DesRoches, 2000). They also receive fewer preventive services and are less likely to have regular care for chronic conditions such as hypertension and diabetes. Chronic diseases can lead to expensive and disabling complications if they are not well managed (Lurie et al., 1984; Lurie et al., 1986; Ayanian et al., 2000). One national survey asked more than 3,400 adults about 15 highly serious or morbid conditions. Of those reporting any such symptoms (16 percent of those surveyed), and with adjustment for demographic and economic characteristics, health status, and a regular source of care,1 an uninsured person was far less likely than someone with insurance to receive care for the reported condition (odds ratio =0.43) (Baker et al., 2000). Additional evidence is presented later in this chapter in the discussion of insurance and access to health care.

Myth: People without health insurance are young and healthy and choose to go without coverage. Almost half (43 percent) of those surveyed in 2000 believed that people without health insurance are more likely to have health problems than people with insurance. About as many (47 percent) thought the likelihood of health problems is about the same for insured and uninsured people (NewsHour-Kaiser, 2000). Voters and policy makers in focus group discussions characterize those without insurance as young people who have the opportunity to be covered and feel they do not need it (Porter Novelli, 2001).

Reality: Compared to those with at least some private coverage, the uninsured are less likely to report being in excellent or very good health (Agency for Healthcare Research and Quality, 2001). In contrast, people reporting excellent or very good health are more likely to be insured. Among those under age 65 who are in fair or poor health, nearly one in five lacks health insurance (Rhoades and Chu, 2000). Of young adults (ages 19–34 years) in poor health, 16 percent are uninsured and 27 percent of those reporting fair health status are uninsured (Figure 1.1) (Agency for Healthcare Research and Quality, 2001).

FIGURE 1.1. Probability of being uninsured for young adults, ages 19 to 34 years, by self-reported health status, 1999.


Probability of being uninsured for young adults, ages 19 to 34 years, by self-reported health status, 1999. SOURCE: Center for Cost and Financing Studies, Agency for Healthcare Research and Quality, based on MEPS data.

Young adults between 19 and 34 are far more likely to lack health insurance than any other age group. This is chiefly because they are less often eligible for employment-based insurance due to the nature of their job or their short tenure in it. They are also more likely than older adults to be in excellent or very good health and consequently may forgo the cost of workplace coverage if it is offered. Turning down a ‘workplace offer is not, however, a significant factor in explaining their lack of coverage. Younger workers accept workplace offers of coverage more often than not, and only 4 percent of all workers between 18 and 44 years of age, roughly 3 million, are uninsured after turning down workplace insurance (Custer and Ketsche, 2000b). Another 11 million uninsured workers between the ages of 18 and 44 (15 percent) hold jobs that do not include an offer of coverage.

The perception that people without insurance have better-than-average health follows from confusing the relatively young age profile of the uninsured with the better health, on average, of younger persons. This obscures the link between health status and health insurance. For those without access to workplace health insurance, poor health is a potential barrier to purchasing nongroup coverage because such coverage may be highly priced, exclude preexisting conditions, or be simply unavailable. Older women (55–64 years) in the work force are especially at risk of being uninsured for this reason: 23 percent of those in good, fair, or poor health have no coverage compared to 10 percent of those in excellent or very good health (Monheit et al., 2001).

Myth: The number of uninsured Americans is not particularly large and has not changed in recent years. Seven out of ten respondents in a nationally representative survey thought that fewer Americans lacked health insurance than actually do (Fronstin, 1998). Roughly half (47 percent) believed that the number of people without health insurance decreased or remained constant over the latter half of the last decade (Blendon et al., 1999).

Reality: During 1999, an estimated 42 million people in the United States lacked health insurance coverage (Mills, 2000). This number represents about 15 percent of a total population of 274 million persons. According to Census Bureau statistics, the number of Americans under age 65 without health insurance grew from 39 million (17 percent of the population under age 65) in 1994 to 44 million (18 percent) in 1998, before falling to 42 million (17 percent) in 1999 (Fronstin, 2000d). This drop of almost 2 million in the number of people ‘without insurance (a reduction of about 4 percent) is certainly a positive change. With a softer economy in 2000 the latest reported gains in insurance coverage may not continue (Fronstin, 2001). The decline in the number of uninsured will not continue if the economy remains slow and health care costs continue to outpace inflation. Due to the lag in measurement and reporting, however, the Census Bureau estimate of health insurance coverage for 2000 may show a further decline in the uninsured rate. This is because the data were collected for a period of strong economic performance.

Of the estimated 42 million people who were uninsured, all but about 420,000 (about 1 percent) were under 65 years of age, the age at which most Americans become eligible for Medicare;2 32 million were adults between ages 18 and 65, about 19 percent of all adults in this age group; and 10 million were children under 18 years of age, about 13.9 percent of all children (Mills, 2000). Throughout this report, the discussion focuses on these uninsured working-age adults and children.

These estimates of the number of persons uninsured are generated from the annual March Supplement to the Current Population Survey (CPS), conducted by the Census Bureau. Unless otherwise noted, national estimates of people without health insurance and proportions of the population with different kinds of coverage are based on the CPS, the most widely used source of estimates of insurance coverage and uninsurance rates. Seven different governmentally and privately sponsored surveys can, however, be used to make nationally representative estimates of the number of people without health insurance. These surveys and the estimates they yield are described briefly in Table B.1 in Appendix B. These surveys differ in size and sampling methods, the questions that are asked about insurance coverage, and the time period over which insurance coverage or uninsurance is measured (Lewis et al., 1998, Fronstin, 2000a). Each survey produces a different estimate of the number of Americans without insurance. The estimates range from 32 million (e.g., Medical Expenditure Panel Survey, 1996, uninsured throughout the year) to 42 million (CPS, 1999, uninsured throughout the year).3

The CPS has been criticized for producing estimates of persons uninsured throughout the year that are too high and probably reflect periods without insurance of less than a full year as well as underreporting of Medicaid coverage (Swartz, 1986; Lewis et al., 1998; Fronstin, 2000a). The Census Bureau has recently revised its survey questionnaire to include an additional question verifying that the respondent means to report lack of coverage over the entire previous year (see Appendix B for further explanation). Still, the CPS is especially useful because it produces annual estimates relatively quickly, reporting the previous year's insurance coverage estimates each September, and because it is the basis for a consistent set of estimates for more than 20 years, allowing for analysis of trends in coverage over time. For these reasons, as well as the extensive use of the CPS in other studies of insurance coverage that are presented in this report, we rely on CPS estimates, with limitations noted. The Committee finds the variation among estimates of the number of persons uninsured produced by the different surveys less critical to its analysis than the order of magnitude of the range of estimates that these surveys yield.

The estimate of the number of uninsured people expands when a population's insurance status is tracked for several years. Over a three-year period beginning early in 1993, 72 million people, 29 percent of the U.S. population, were without coverage for at least one month. Within a single year (1994), 53 million people experienced at least a month without coverage (Bennefield, 1998a).

Myth: Most people who lack health insurance are in nonworking families. An April 2000 national telephone survey by the NewsHour with Jim Lehrer-Kaiser Family Foundation found that 57 percent of the adults polled believed that most people without health insurance were unemployed or from families with unemployed adults (News Hour—Kaiser, 2000). Other surveys report comparable findings (Blendon et al., 1999; Wirthlin Worldwide, 2001).

Reality: More than 80 percent of uninsured children and adults under the age of 65 live in working families. Six out of every ten uninsured adults are themselves employed. Although working does improve the likelihood that one and one's family members will have insurance, it is not a guarantee. Even members of families with two full-time wage earners have almost a one-in-ten chance of being uninsured (9.1 percent uninsured rate) (Hoffman and Pohl, 2000). See Chapter 3, especially Figures 3.1 and 3.2, for further details.

Myth: New immigrants account for a substantial proportion of people without health insurance. One analysis has attributed a significant portion of the recent growth in the size of the U.S. uninsured population to immigrants who arrived in the country between 1994 and 1998 (Camarota and Edwards, 2000).

Reality: Recent immigrants (those who came to the United States within the past four years) do have a high rate of being uninsured (46 percent), but they and their children account for just 6 percent of those without insurance nationally (Holahan et al., 2001). In fact, there has been a net decrease in the number of recently arrived immigrants since 1994 (Holahan et al., 2001). Overall, noncitizens account for fewer than one in five uninsured persons (Mills, 2000).

Myths and Policy Making

Popular confusion about the facts of health insurance coverage and its importance can hamper effective policy making, as can policy makers' uncertainty about the interpretation of coverage trends and consequences. This report and those that will follow aim to provide reliable information, useful to both the public and policy leaders—legislators, employers, program managers—as they meet the ongoing challenges of financing health care.


Measuring Impacts of Coverage

Health insurance coverage is a key element in most models that depict access to health care. The relationship between health insurance and access to care is well established, as documented later in this chapter. Although the relationship between health insurance and health outcomes is neither direct nor simple, an extensive clinical and health services research literature links health insurance coverage to improved access to care, better quality, and improved personal and population health status. The Committee's conceptual framework for considering the extent and nature of these and additional effects of health insurance builds selectively upon the most widely used behavioral model of access to health services (Andersen, 1995; Andersen and Davidson 2001). The framework focuses primarily on the economic, financial, and coverage-related factors that facilitate the use of health care services. The Committee uses the framework in this introductory report to conceptualize various effects of health insurance and to provide an overview of the subsequent analyses in future reports (see Figure ES.2 and Appendix A for a further description of this model).

The Committee will use this conceptual model to identify, organize, and assess the evidence regarding important consequences of uninsurance, each of which will be the subject of a future report: individual health outcomes, family well-being, community impacts, and economic costs for society as a whole. Figure 1.2 depicts the relationship among the topics of the Committee's reports in terms of a series of overlapping circles. For example, the second report, on personal health outcomes for uninsured adults, is represented by the innermost circle of the figure, while the third report, on family well-being, encompasses the subjects of the second report but emphasizes a different unit of analysis, namely, the family. The sixth report in the series will present information about strategies and initiatives undertaken locally, statewide, or nationally to address the lack of insurance and its adverse impacts. Each of these planned reports is described briefly in Chapter 4.

FIGURE 1.2. Levels of analysis for examining the effects of uninsurance.


Levels of analysis for examining the effects of uninsurance.

Scope of This Report

This discussion of health insurance coverage focuses primarily on the U.S. population under age 65 because virtually all Americans 65 and older have Medicare or other public coverage. Furthermore, it focuses specifically on those without any health insurance for any length of time. While the effects of lacking health insurance on access to care and thus potentially on health may not be apparent for those who are uninsured only briefly, even short periods without insurance entail a measure of financial risk to self and family of incurring high expenses for health care.

The Committee does not attempt to address the condition of “underinsurance.” By the “underinsured” is meant individuals or families whose health insurance policy or benefits plan offers less than adequate coverage. Most people would consider themselves underinsured if their health plan required extensive out-of-pocket payments in the form of deductibles, coinsurance or copayments, or maximum benefit limits. Many policies also exclude specific services such as mental health treatment, long-term care, or prescription drugs. The problems faced by the underinsured are in some respects similar to those faced by the uninsured, although they are generally less severe. Uninsurance and underinsurance, however, involve distinctly different policy issues, and the strategies for addressing them may differ. Throughout this study and the five reports to follow, the main focus is on persons with no health insurance and thus no assistance in paying for health care beyond what is available through charity and safety net institutions.


For individuals and families, health insurance both enhances access to health services and offers financial protection against high expenses that are relatively unlikely to be incurred as well as those that are more modest but are still not affordable to some. Health insurance is a powerful factor affecting receipt of care because both patients and physicians respond to the out-of-pocket price of services. Health insurance, however, is neither necessary nor sufficient to gain access to medical services. Nonetheless, the independent and direct effect of health insurance coverage on access to health services is well established. This section documents that research literature and presents the Committee's findings regarding access to care.

Subsequent Committee reports will build on this finding and evaluate evidence for the further relationship between insurance coverage and health outcomes. Appendix A describes and depicts schematically the Committee's conceptual model of this complex relationship, which is affected by a variety of personal, economic, and social factors and health care processes that are in turn subject to many influences.

Health Insurance Facilitates Access to Care

Many people who lack health insurance will forgo the care they need until their condition becomes intolerable. Others will obtain the health care they need even without health insurance, by paying for it out of pocket or seeking it from providers who offer care free or at highly subsidized rates. For still others, health insurance alone does not ensure receipt of care because of other nonfinancial barriers, such as a lack of health care providers in their community, limited access to transportation, illiteracy, or linguistic and cultural differences. Nonetheless, health insurance remains a key factor in assuring access to health care.

Formal research about uninsured populations in the United States dates to the late 1920s and early 1930s when the Committee on the Cost of Medical Care produced a series of reports about financing physician office visits and hospitalizations. This issue became salient as the numbers of medically indigent climbed during the Great Depression. With the rise of commercial insurers and the decline of community rating offered by Blue Cross-Blue Shield and other nonprofit insurers in the 1950s, new studies of individual and family health expenditures were co-sponsored by the University of Chicago and the Health Information Foundation. These studies became the factual basis for legislation that was enacted as the Medicare and Medicaid amendments to the Social Security Act in 1965.

Since the enactment of Medicare and Medicaid, health services research on uninsured populations has been sponsored federally and privately, at increasing levels of support over time and using new survey tools and data sets (Somers and Somers, 1961; Numbers, 1979; Starr, 1982; Andersen and Anderson, 1999). The Census Bureau started collecting detailed information about health insurance in the latter half of the 1970s and the National Center for Health Services Research, a predecessor to AHRQ, conducted the National Medical Care Expenditure Survey (NMCES) in 1977, followed by the National Medical Expenditure Survey (NMES) in 1987 and AHRQ's Medical Expenditure Panel Survey (MEPS), launched and conducted annually since 1996. A summary of the major surveys collecting health insurance and utilization information is presented in Appendix B.

Population-based surveys have been used to examine access to health services by measuring components of primary care, such as number of physician visits and immunization rates, sites of care (e.g., physician office, hospital outpatient department, clinic), barriers to care (e.g., inability to pay), and unmet health needs (e.g., health status, inability to obtain care when needed) (Andersen and Aday, 1978; Aday et al., 1984; Lurie et al., 1984; Monheit et al., 1985; Lurie et al., 1986; Hafner-Eaton, 1993; Newacheck et al., 1993; Himmelstein and Woolhandler, 1995; Sox et al., 1998; Hsia et al., 2000; Kasper et al., 2000). The likelihood of having any physician visit within a year, the number of visits annually, and having a regular source of care are well established measures of access. Empirical studies consistently support the link between access to care and improved health outcomes (Bindman et al., 1995; Starfield, 1995).

Having a regular source of care can be considered a predictor of access, rather than a direct measure of it, when health outcomes are themselves used as access indicators. This extension of the notion of access measurement was made by the IOM Committee on Monitoring Access to Personal Health Care Services (Millman, 1993, p.33):

“[T]he committee defined access as follows: the timely use of personal health services to achieve the best possible health outcomes. Importantly, this definition relies on both the use of health services and health outcomes to provide yardsticks for judging whether access has been achieved.”

Thus, in Access to Health Care in America, the earlier IOM committee incorporated health outcomes into the definition of access. In this first report of the Committee on the Consequences of Uninsurance, consideration of the relationship between health insurance and access is limited to well established findings regarding process measures of access. The next report, which will examine health outcomes for the uninsured, will evaluate clinical and epidemiological research evidence in terms of the more demanding concept of realized access to health care.

The likelihood that those without health insurance lack a regular source of care has increased substantially since 1977. In 1996, people without insurance were 2.5 times more likely to lack a regular source of care than were the insured (Zuvekas and Weinick, 1999; Weinick et al., 2000). Children without insurance were three times as likely as children with Medicaid coverage to have no regular source of care (15 percent versus 5 percent), and uninsured adults were more than three times as likely as either privately or publicly insured adults to lack a regular source of care (35 percent versus 11 percent) (Haley and Zuckerman, 2000).

The benefits for children of having health insurance and a regular source of care, in terms of routine physician visits and appropriate preventive care, are well documented (Lave et al., 1998; Newacheck et al., 1998; Haley and Zuckerman, 2000). However, the impact of parents' health and health insurance on the well-being of their children has received attention only recently. Whether or not parents are insured appears to affect whether or not their children receive care— as well as how much care—even if the children themselves have coverage (Hanson, 1998). The health of parents can affect their ability to care for their children and the level of family stress. Worrying about their children's access to care is itself a source of stress for parents.

Uninsured adults are less likely to receive health services, even for certain serious conditions. In a study described earlier (Baker et al., 2000), even after adjusting for differences in age, sex, income, and health status, uninsured people were less than half as likely as insured persons to receive care for a condition that physicians deemed highly serious and requiring medical attention. People without insurance are also less likely than people with insurance to receive preventive services and appropriate routine care for chronic conditions, even as the importance of preventive care and the prevalence of chronic disease become more prominent elements within health care (Hafner-Eaton, 1993; Ayanian et al., 2000; Institute of Medicine, 2001). Finally, those who lack health insurance are more likely to be hospitalized for conditions that might have been avoided with timely ambulatory care (Weissman et al., 1992; Kozak et al., 2001).

The level of out-of-pocket costs for care has been demonstrated in randomized trials, natural experiments, and observational studies to have substantial effects on the use of health care services (Newhouse et al., 1993; Zweifel and Manning, 2000). Table 1.1 gives a sense of the magnitude of these costs. Uninsured patients may be charged more than patients with coverage, who benefit from discounts negotiated by their insurer, which amplifies the financial impact of lacking coverage (Wielawski, 2000; Kolata, 2001).

TABLE 1.1. Illustrative Charges to Patients, Insured and Uninsured, 1999 (in dollars).


Illustrative Charges to Patients, Insured and Uninsured, 1999 (in dollars).

Differential Access to Care for the Uninsured

Not only do persons without insurance receive less care, but the providers who serve them differ systematically from those who treat insured patients. Public hospitals, health departments, and health clinics (e.g., community, migrant, or rural health centers) are more likely than other providers to serve uninsured persons, two-thirds of whom are members of lower-income families (annual income below 200 percent of the federal poverty level [FPL]: $33,400 for a family of four in 1999). These institutions generally receive public funding to support the provision of free or reduced fee care to those who cannot afford to pay private fees. They serve as “core safety-net providers,” with two distinguishing characteristics:

“(1) either by legal mandate or explicitly adopted mission they maintain an “open door,” offering access to services for patients regardless of their ability to pay; and (2) a substantial share of their patient mix is uninsured, Medicaid, and other vulnerable patients” (Institute of Medicine, 2000).

The IOM report on the safety net also stressed the diversity of local safety-net providers and services across states, communities, and geographic regions. In rural areas, for instance, the mix of safety-net providers tends to feature private physicians and health centers or clinics, whereas urban areas are more likely to be served by teaching hospitals (Schur and Franco, 1999).

In addition to those providers whose patient populations include substantial proportions of uninsured persons, in the aggregate, private physicians, community hospitals, and teaching hospitals affiliated with academic health centers provide significant amounts of care to uninsured patients (Cunningham and Tu, 1997; Mann et al., 1997; Institute of Medicine, 2000). Nationally representative surveys show that between two-thirds and three-quarters of physicians report providing some charity care, accounting for about 5 percent of their case load on average (Foreman, 1992; Cunningham, 1999b).

The wide geographic variation in the organization, financing, and delivery of health services contributes to the scarcity of quantitative information about services for uninsured people. Compared with insured persons, greater numbers of uninsured persons obtain care from hospitals and clinics or health centers than from office-based physicians, and are less likely to identify a person, rather than a facility, as their usual source of care (Shi, 2000a). Uninsured patients are less than half as likely as insured patients to report that a physician's office is their usual source of care (about one-third of all uninsured patients compared with about two-thirds of the general population) (Cunningham and Whitmore, 1998).

Hospital emergency departments or outpatient departments serve as the regular source of care for one out of every six uninsured patients that report having a regular source of care (Weinick et al., 1997). A substantial proportion of emergency department visits is for nonurgent conditions (Pane et al., 1991; Grumbach et al., 1993; Baker et al., 1994; Zimmerman et al., 1996). Because hospital emergency departments are legally required to assess and stabilize all patients with any medical condition without regard for ability to pay, they are the only providers who cannot turn uninsured patients away for lack of a source of payment.4 Although emergency departments are portrayed as a costly and inappropriate site of primary care services, many uninsured patients seek care in emergency departments because they are sent there by other health care providers or have nowhere else to go. Emergency care specialists argue that the nation's emergency departments not only serve as providers of last resort but are a critical entry point into the health care system (O'Brien et al., 1999).


Three chapters follow in this report. Chapter 2 provides an overview of how employment-based health insurance, public programs and individual insurance policies operate and interact to provide extensive but incomplete coverage of the U.S. population. This includes a review of historical trends and public policies affecting both public and private insurance, a discussion of the interactions among the different types of insurance, and an examination of why people move from one program to another or end up with no coverage.

Chapter 3 synthesizes existing information to arrive at a composite description of the uninsured: What characteristics do people without coverage often share? Where do the uninsured live? The chapter also presents information about the risk of being or becoming uninsured: How does the chance of being uninsured change depending on selected characteristics, such as racial and ethnic identity, rural or urban residency, and age? What are the probabilities for specific populations, such as racial and ethnic minorities, rural residents, and older working-age persons, of being uninsured? How does the chance of being uninsured change over a lifetime?

In addition to characterizing the likelihood of being uninsured in terms of a single dimension, such as gender, age, race, work status, or geographic region, Chapter 3 also presents the results of multivariate analyses that offer a more informative depiction of the factors that contribute to the chances of being uninsured.

Finally, in Chapter 4 the Committee presents the research agenda for its overall project and previews the five future reports.

Box 2.1

In the United States, health insurance is a voluntary matter, yet many people are involuntarily without coverage. There is no guarantee for most people under the age of 65 that they will be eligible for or able to afford to purchase or retain health insurance.

  • Almost seven out of every ten Americans* under age 65 years are covered by employment-based health insurance, either from their job or through a parent or spouse. Three quarters of workers are offered health insurance by their employers, and most decide to purchase or take up the offer of coverage. Of the 17 percent of workers who decline an employer's offer, about a quarter, or 4 percent of workers overall, remain uninsured.
  • Individually purchased policies and public insurance (primarily Medicaid) together cover one out of five persons under age 65.* Both have limitations. Poor health status or low income may preclude the purchase of an affordable individual policy. The combination of strict eligibility requirements and complex enrollment procedures makes public coverage often difficult to obtain and even more difficult to maintain over time.
  • A change in insurance premium or terms, as well as changes in income, health, marital status, terms of employment, or public policies, can trigger a loss or gain of health insurance coverage. For about one-third of the uninsured population, being without coverage is a temporary or one-time interruption of coverage, and the median duration of a period without insurance is between 5 and 6 months. Uninsured persons in low-income families and those with less education experience longer periods without coverage, on average, than their higher income and more educated counterparts.
  • Insurance industry underwriting practices, the costs of health services, and the patchwork of public policies regarding insurance coverage all contribute to the economic pressures on employers, insurers, and government programs offering health insurance. Small firms are especially likely to face high costs. Workers who take up an employer's offer of a subsidized health benefit typically pay directly between one-quarter and one-third of the total cost of their insurance premium, in addition to paying deductibles, copayments, and the costs of health services that are not covered or are covered only in part by their health plan. For families earning less than 200 percent of the federal poverty level, these expenses can exceed 10 percent of their annual income.
  • Since the mid-1970s, growth in the cost of health insurance has outpaced the rise in real income, creating a gap in purchasing ability that has added roughly one million persons to the ranks of the uninsured each year. Despite the economic prosperity of recent years, between 1998 and 1999 there was only a slight drop in the numbers and proportion of uninsured Americans. Through the early 1990s, the rising uninsured rate reflected a decline in employment-based coverage. Since the mid-1990s, increases in employment-based coverage have been offset by steady or declining rates of public and individually purchased coverage.

Altogether, about 83 percent of the nonelderly population is covered by employment-based, individual and public plans. Some people report more than one source of coverage over the course of a year.



“Regular source of care” is defined as the place or provider from which one usually seeks care or advice about health care. A regular source of care may be a physician's office, a clinic, a health plan facility or a hospital emergency room or outpatient clinic. Optimally, one's regular source of care provides continuity of attention, facilitates access to appropriate services, and maintains records.


Medicare, the federal insurance program for the elderly, disabled, and those with end-stage renal disease, provides almost universal coverage for hospital care for those over age 65. A small fraction of the elderly do not qualify for the program because they do not have sufficient Social Security work credits. The clergy and other religious workers comprise the largest single category of people without ties to Social Security and Medicare.


In 1996, the CPS estimate of the number of nonelderly persons uninsured was 41 million (Fronstin, 2000a).


The federal Emergency Medical Treatment and Active Labor Act, part of the Consolidated Omnibus Budget Reconciliation Act of 1985, requires hospital emergency rooms to assess and stabilize all patients with a life- or limb-threatening or emergency medical condition or those who are about to give birth. Hospitals are not required to provide continuing care after the patient has been stabilized and transferred or released. No federal funds directly support this mandate.

Image p20004966g61001
Image p20004966g62001
Image p20004966g7001
Copyright 2001 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK223643


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...